Productivity PerspectivesProductivity Perspectivesdepend on your point of viewdepend on your point of view

Eric BartelsmanEric Bartelsman

Vrije Universiteit Amsterdam and Tinbergen Vrije Universiteit Amsterdam and Tinbergen InstituteInstitute

Canberra, ABS/PC Dec. 9, 2004Canberra, ABS/PC Dec. 9, 2004

OverviewOverview

Recent Evidence on ProductivityRecent Evidence on ProductivityWho Cares about Productivity? Why?Who Cares about Productivity? Why?Integrating the EvidenceIntegrating the EvidenceWhat do we really know? (What will we tell What do we really know? (What will we tell

policy makers?)policy makers?)Research AgendaResearch Agenda

Recent EvidenceRecent Evidence

Growth AccountingGrowth Accounting Contribution of ICT, other capital, MFPContribution of ICT, other capital, MFP Sectoral, cross-country, timeseriesSectoral, cross-country, timeseries

Econometric EstimatesEconometric Estimates Cross-country convergence (million regressions)Cross-country convergence (million regressions) Sectoral and micro-level datasetsSectoral and micro-level datasets

Growth AccountingGrowth Accounting

US EU NL AUS

LaborProductivityGrowth

1.9 1.4 1.1 3.2

ICT .7 .4 .6 1.2

OtherCapital

.3 .5 .1 .4

TFP .8 .5 .5 1.6Source: van Ark & O’Mahony, Parham (AUS)

1995-2000: contribution in pct-points to average annual growth

Growth AccountingGrowth Accounting

Only factors that are explicitly purchased can Only factors that are explicitly purchased can contribute to output in the frameworkcontribute to output in the framework

no theory for role of spillovers or policy environmentno theory for role of spillovers or policy environment

Representative firm makes optimizing decisionsRepresentative firm makes optimizing decisions perfect competition is assumedperfect competition is assumed

Contribution of quality of capital or labor can be Contribution of quality of capital or labor can be computedcomputed

no account of no account of sourcesource of higher quality of higher quality

But: most of output growth is accounted forBut: most of output growth is accounted for

Sectoral ComparisonsSectoral Comparisons

US EU NL

LaborProductivity

2.2 .7 .8

ICT-Producing

.6 .5 .2

ICT-Using 1.3 .3 .3

Other .3 .0 .3

Source: van Ark & O’Mahony

1995-2002: contribution in pct-points to average annual growth

Convergence in LevelsConvergence in Levels

GDP per hour

0.20

0.40

0.60

0.80

1.00

1.20

19641967197019731976197919821985198819911994199720002003

rela

tive

to

US

A

AUS FRA JAP NLD UK

Source: GGDC, Groningen

ConvergenceConvergence

What drives convergence process?What drives convergence process? Many factors are significant in cross-country growth Many factors are significant in cross-country growth

regressionsregressions

Do we really know if 10% above or below is Do we really know if 10% above or below is significantsignificant

international price comparisonsinternational price comparisons definitions of output and inputdefinitions of output and input

Sectoral and micro dataSectoral and micro data

Econometric evidence in panel data settingEconometric evidence in panel data setting R&DR&D

R&D output elasticitiesR&D output elasticitiesShifting of R&D based on tax-incentives Shifting of R&D based on tax-incentives Spillover benefits of R&D done abroadSpillover benefits of R&D done abroad

ICTICTFirm-level evidence on return to ICT investmentFirm-level evidence on return to ICT investment

Many other studies: trade-openess, competition, schoolingMany other studies: trade-openess, competition, schooling

OverviewOverview

Recent Evidence on ProductivityRecent Evidence on ProductivityWho Cares about Productivity? Why?Who Cares about Productivity? Why?Integrating the EvidenceIntegrating the EvidenceWhat do we really know? (What will we tell What do we really know? (What will we tell

policy makers?)policy makers?)Research AgendaResearch Agenda

Who cares about Productivity?Who cares about Productivity?

Central BanksCentral Banks productivity levelsproductivity levels

used for computing potential GDPused for computing potential GDP productivity growth ratesproductivity growth rates

monitor cycle. Used in hours-to-output modelsmonitor cycle. Used in hours-to-output models

Economic Policy MakersEconomic Policy Makers Comparisons across countries: measures of wellbeingComparisons across countries: measures of wellbeing Comparisons across countries/over time: targets for policy Comparisons across countries/over time: targets for policy

ambition (ie EU ‘Lisbon’)ambition (ie EU ‘Lisbon’) Extremely useful for Extremely useful for Policy EvaluationPolicy Evaluation

Why?Why?

Innovation/R&D policy: how much to spend, where, Innovation/R&D policy: how much to spend, where, and in what manner.and in what manner.

Education: total budget and allocation across types of Education: total budget and allocation across types of students/workersstudents/workers

Trade: effects of imported materials; effects of Trade: effects of imported materials; effects of international specialisationinternational specialisation

Competition policy: effects of concentration; Competition policy: effects of concentration; deregulationderegulation

Framework conditions: which regulations to tackle Framework conditions: which regulations to tackle first; how do they affect productivity first; how do they affect productivity

OverviewOverview

Recent Evidence on ProductivityRecent Evidence on ProductivityWho Cares about Productivity? Why?Who Cares about Productivity? Why?Integrating the EvidenceIntegrating the EvidenceWhat do we really know? (What will we tell What do we really know? (What will we tell

policy makers?)policy makers?)Research AgendaResearch Agenda

Integrating Evidence on Integrating Evidence on ProductivityProductivity

Firms choose innovation strategyFirms choose innovation strategy– Technology frontier populated by innovating firmsTechnology frontier populated by innovating firms

– Adopters buy latest technology, and apply it wellAdopters buy latest technology, and apply it well

– Followers wait until capital can be used ‘off-the-shelf’Followers wait until capital can be used ‘off-the-shelf’

Firms then choose output/inputs to meet Firms then choose output/inputs to meet demanddemand

Innovate, Adopt or FollowInnovate, Adopt or Follow

Innovation is highly risky and requires Innovation is highly risky and requires knowledge and risk-funding. But, it also knowledge and risk-funding. But, it also requires flexibility in resources to leverage requires flexibility in resources to leverage successful ideas. Further, markets need to successful ideas. Further, markets need to respond to innovative productsrespond to innovative products

Adopters need some skilled workers and Adopters need some skilled workers and functioning financial markets.functioning financial markets.

Innovatevs Adopt

Adopt vsFollow

Cost of InnovativeActivity R&D, Skills X X

Foreign R&D, VC X

Reallocation

Flexibility in labor andproduct markets

X

Financial Markets X X

Choice of resourcesChoice of resources

After choosing innovation strategy, firms After choosing innovation strategy, firms attempt to meet resulting market demandattempt to meet resulting market demand

At firm level, growth accounting is valid toolAt firm level, growth accounting is valid tool– innovators push out frontier, if successfulinnovators push out frontier, if successful– adopters have embodied technical changeadopters have embodied technical change– followers ‘converge’ towards previous frontierfollowers ‘converge’ towards previous frontier

Aggregate productivity is weighted average Aggregate productivity is weighted average of firms’ productivitiesof firms’ productivities

OverviewOverview

Recent Evidence on ProductivityRecent Evidence on ProductivityWho Cares about Productivity? Why?Who Cares about Productivity? Why?Integrating the EvidenceIntegrating the EvidenceWhat do we really know? (What will we tell What do we really know? (What will we tell

policy makers?)policy makers?)Research AgendaResearch Agenda

What to tell policymakers?What to tell policymakers?

What to tell policymakers?What to tell policymakers?

Productivity growth requires:Productivity growth requires:– human capitalhuman capital

– flexibility in resourcesflexibility in resources

– competitioncompetition

Productivity and innovation are harmed by:Productivity and innovation are harmed by:– targeted policies to firms/sectors/groupstargeted policies to firms/sectors/groups

– emphasis on ‘hard sciences’emphasis on ‘hard sciences’

– squeezing fundamental researchsqueezing fundamental research

OverviewOverview

Recent Evidence on ProductivityRecent Evidence on ProductivityWho Cares about Productivity? Why?Who Cares about Productivity? Why?Integrating the EvidenceIntegrating the EvidenceWhat do we really know? (What will we tell What do we really know? (What will we tell

policy makers?)policy makers?)Research AgendaResearch Agenda

Research AgendaResearch Agenda

TheoryTheory– models of heterogeneous firms in dynamic market models of heterogeneous firms in dynamic market

(Acemoglu, Klette&Kortum, Melitz&Helpman, Aghion (Acemoglu, Klette&Kortum, Melitz&Helpman, Aghion et al)et al)

StatisticsStatistics– longitudinal business datalongitudinal business data

– linked employer-employee datalinked employer-employee data

– linked ‘special surveys’/policy experimentslinked ‘special surveys’/policy experiments

– international distributed micro-data analysisinternational distributed micro-data analysis

Provision of metadata.Approval of access.Disclosure analysis

of cross-country tables.Disclosure analysis of Publication

Res

earc

her

Policy QuestionResearch Design Program Code

Publication

Net

wor

k Metadata

Networkmembers

Cross-countryTables

NS

Os

Distributed micro data research